hastrup, kirsten - getting it right

http://ant.sagepub.com/

Anthropological Theory

http://ant.sagepub.com/content/4/4/455The online version of this article can be found at:

DOI: 10.1177/1463499604047921

2004 4: 455Anthropological TheoryKirsten Hastrup

Getting it right : Knowledge and evidence in anthropology

Published by:

http://www.sagepublications.com

can be found at:Anthropological TheoryAdditional services and information for

http://ant.sagepub.com/cgi/alertsEmail Alerts:

http://ant.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://ant.sagepub.com/content/4/4/455.refs.htmlCitations:

at Copenhagen University Library on September 7, 2010ant.sagepub.comDownloaded from


http://ant.sagepub.com/content/4/4/455

http://www.sagepublications.com

http://ant.sagepub.com/cgi/alerts

http://ant.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://ant.sagepub.com/content/4/4/455.refs.html


Anthropological Theory

Copyright © 2004 SAGE Publications(London, Thousand Oaks, CA and New Delhi)

www.sagepublications.comVol 4(4): 455–472

10.1177/1463499604047921

455

Getting it rightKnowledge and evidence in anthropology

Kirsten HastrupUniversity of Copenhagen, Denmark

AbstractIn this article the nature of anthropological knowledge is discussed with a view to areassessment of the call for evidence. It will be argued that the traditional view ofevidence as being somehow outside a particular argument is untenable, given the wayin which anthropology accesses knowledge by engaging a particular field. This may beseen to lead to an uneasy question of authority in anthropology, yet at closerinspection, it is possible to establish new grounds for anthropological authority and toarrive at a new sense of ‘getting it right’. The general discussion is substantiated byreference to the author’s work on Icelandic history as well as her fieldwork in Iceland.

Key Wordsauthority • causality • epistemology • ethics • experience • explanation • knowledge

The question of evidence is acute if anthropology shall aspire to anything but reportingquaint stories from strange places. In a post-positivist era, however, I shall argue that itis not possible to adhere to old notions of ‘evidence’ as external to the context of thesituation. In that vein, the ambition of this article is to contribute to a new awarenessof the anthropological mode of knowing about the world and of ‘getting it right’. I shallstart by briefly identifying what we may mean by knowledge in this day and age, as aprecursor to a discussion of the traditional positivity of evidence in the pursuit of expla-nation. To substantiate these general discussions, I shall investigate, first, the nature ofhistorical explanation – or the identification of causes in history – with particular refer-ence to my own work on Icelandic history. Next, I shall inquire into the making ofconnections in the field – or the explanation of social phenomena; again it will be brieflylinked up to my own work in Iceland. I terminate by some general remarks on how toget it ‘right’, and how this implicates us deeply in a narrative ethics upon which theanthropological authority rests.

KNOWLEDGE AND THE MODE OF KNOWINGIf we concede to the proposition that anthropology is a distinct field of knowledge, it isworth looking closer at the notion of knowledge itself. What seems clear is that

04 047921 hastru (jr/t) 5/11/04 8:44 am Page 455


www.sagepublications.com


knowledge must be organized information; in the case of anthropology it concerns theorganized information about ways of living in the world and modes of attending to theworld. The organization implies that knowledge is both reductive and selective (Schmidt,1991: 17). It is reductive because it renders empirical complexity and messiness in clear,but therefore also more limited, propositions about the world. It is selective, because forit to be knowledge it has to disregard some information. To investigate a particular ideaof knowledge, we must, therefore, not only look at the object of interest but also themode of that interest, that is the particular way of attending to the object and of organiz-ing the information as knowledge. If anthropologists are interested in cultures (orsocieties), they are so in a particularly anthropological way – learnt explicitly throughtraining and more implicitly by way of exemplars. If we make no claim to such distinc-tion, we cannot claim to have a particular field of scholarship. This implies that knowingis a matter of perspective; there is no knowledge without someone who knows in aparticular way. Knowledge, therefore, is a social phenomenon rather than simply asubstance. To maintain scholarly authority one must be able to account for the particu-lar mode of interest that gives direction and shape to knowledge.

In anthropology the mode of knowing is intimately connected to the tradition offieldwork; a lot has been said about this, and I just want to emphasize that through theeffort at getting into ‘another’ world (defined as ‘other’ for the purpose of analysis), theanthropologist learns something new about it; effort is the operative word here. I shallreturn to this. Here it suffices to say that because of its particular method, anthropo-logical knowledge is defined as much by an epistemology as by the purported ontologyof its object of interest. This leads us further into the discussion of what we should meanby knowledge in the first place.

In the modernist era, anthropological knowledge was presented as knowledge aboutother cultures; it consisted in largely ontological propositions about the organization of(other) social systems and thoughts. The result was an encyclopaedic knowledge thatposited itself as an object-knowledge – in the triple sense of attaching itself to objects,working by way of objectification, and itself becoming an object to be possessed andrecycled. Gradually, this view outlived itself, because it was realized that most of whathad passed for ontology in anthropology was in fact located in our experience of it, andin the way in which it was registered – or silenced. In consequence, knowledge hasbecome – and must be – acknowledged (implicitly, at least) as relational, both in thesense that it attaches itself to relations between people or between people and objectsand in the sense that it emerges within a dialogical field. If relational knowledge is moreimplicit and ephemeral than object-knowledge, it may nevertheless transform into thelatter with time, partly through the general process of objectification that goes alongwith classification and articulation (also known as ‘ontological dumping’), partlythrough institutional endorsement.

We arrive at a point where we may see how anthropological knowledge not onlyreflects a particular mode of knowing (an epistemology) but incorporates it in a verydirect way. If in fieldwork the anthropologist gains knowledge by way of social relations,this relational aspect has a general bearing on the processes by which facts are establishedas (relevant) facts in the first place. The relation between the ‘knower’ and the ‘object’of necessity bends back into the perception of the object itself and is cemented in writing.As Tim Jenkins has it: ‘the accounts build in the relation of outside observer to object,

ANTHROPOLOGICAL THEORY 4(4)

456




as if it were a property of the object itself ’ (Jenkins, 1994: 443). Knowledge, then,depends intimately on the modes of knowing and of interpreting – both of which aredisciplined. As Malcolm Crick once said: ‘Knowledge is a social achievement: it consistsof meanings that have “made it” ’ (Crick, 1982: 28). Not only have they made it throughthe registration filter of the ethnographer in the field, they have also made it throughthe institutional filters of the academic discipline. Knowledge, therefore, is no simple‘object’, because it bears all the marks of its institution, including a particular ‘style ofreasoning’ that by itself becomes a standard of objectivity (Hacking, 1992: 13).

Anthropologists have always been interested in the question of knowledge, and havecontributed to cross-disciplinary investigations into the relativity of knowledge andrationality (Wilson, 1974; Hollis and Lukes, 1982). While instigated by ethnographicanalyses such as Evans-Pritchard’s of Zande witchcraft (1936), the theoretical discussionof ‘knowledge’ also had important precursors, for instance in Durkheim’s extensivediscussion of the relationship between science and belief (Durkheim, 1915) and in Lévi-Strauss’ work, marking the pinnacle of French rationalism in anthropology and compris-ing a discussion of the contrastive qualities of science and magic (Lévi-Strauss, 1962).The common subtext of these discussions is the relationship between ‘our’ and ‘their’knowledge that at first sight seems obsolete today, but they also point to a still validrecognition of the fact that in human life, there is always both something that we knowand something that we simply sense or feel.

Recently, Fredrik Barth has launched ‘an anthropology of knowledge’ in which hesuggests that knowledge is what a person uses to interpret and to act in the world (2002:1). Knowledge in this sense is purely instrumental, and is hardly distinguishable fromculture, even though Barth suggests that while knowledge gives material for reflectionand premises for action, culture also embraces the results of reflection and action.Another way of distinguishing between knowledge and culture, according to Barth, isthat while culture is equally distributed in a society, knowledge is differentiated amongpeople, partly because they have different experiences. While one can easily see howknowledge is differentiated, it is more difficult (in my view) to accept that ‘culture’ isthe same for everyone. It certainly presupposes a rather anachronistic concept of cultureas an essence that is installed in individuals in equal measure by birth and belonging. Bethat as it may, Barth’s main message is that knowledge can be studied empirically, andin spite of Barth’s ambition to break away from the encyclopaedic concept of knowl-edge, he largely confirms it as far as ‘other people’s’ knowledge is concerned. Knowledgeis empirical matter that can be identified as particular ‘knowledge traditions’ by theobservant ethnographer.

The focus of what people know not only blurs the question of how they know it, butalso tends to overlook the issue of what can be known under particular historical circum-stances. When Barth speaks about (empirically) differentiated knowledge as a corollaryto different experiences, he fails to acknowledge the deeper processes by which differ-entiation of knowledge takes place in all social worlds. Pierre Bourdieu gives a clue:

The process of differentiation of the social world which leads to the existence ofautonomous fields concerns both being and knowledge. In differentiating itself, thesocial world produces differentiation of the modes of knowledge of the world. Toeach of these fields there corresponds a fundamental point of view on the world which

HASTRUP Getting it right

457

04 047921 hastru (jr/t) 5/11/04 8:44 am 57



creates its own object and finds in itself the principle of understanding and expla-nation appropriate to that object. (Bourdieu, 2000: 99)

While, superficially, there may seem to be an accord between Barth’s traditions of knowl-edge and Bourdieu’s differentiated fields of knowledge, the latter are distinguished bythe implicit epistemological awareness about the differentiated fields creating their ownobjects of knowledge. In consequence of this, the empirical notion of (object) knowl-edge must subside to a more theoretical notion of knowledge as primarily a socialphenomenon and only by derivation a corpus, or – in the terms used earlier – a relationfirst and only then (possibly) an object.

This vital shift in the view of knowledge has important implications for the general-izations that we may claim to be ‘right’. Instead of the ‘horizontal’ generalization aboutculture (from wall-to-wall, so to say), anthropologists increasingly move towards‘vertical’ generalizations about the processes by which meanings and practices becometemporarily objectified in social practice, as knowledge or another kind of certainty. Thisimplies that the anthropological object is emergent. The object has no fixed ontologicalstatus, be it as a culture, society, community – or, indeed, knowledge. Evidently, whenit comes to analysis and writing, a sense of closure must be attained; the network mustbe ‘cut’, so to say (Strathern, 1996), in my terms implying a temporary objectificationof relational knowledge, from which others may then proceed – provided they aresatisfied about the soundness of the argument. This takes us to the question of evidence.

THE POSITIVITY OF EVIDENCEMost anthropologists would agree that anthropological knowledge is based in empiricalrealities; anthropology is about the real world. Were we to think about this propositionfor just an instant, uncertainties multiply; what do we man by ‘based in’ empirical reality,and ‘about’ the world. These prepositional terms indicate particular relationships thatare neither self-evident nor neutral, given the suggestion mentioned earlier that therelation to the object bends back into the object itself.

Let us take the simple statement a bit further, however. If anthropology is about theworld, a long tradition of anthropological scholarship has now taught us that this worldis fragmented, fluctuating and endlessly shifting, partly due to individual actions andunprecedented events, partly due to englobing processes of which agents are only dimlyaware – if at all. If the world is shifting, this applies – by definition – also to any science‘about’ it, and the (postmodernist) view of anthropological knowledge as partial andpositioned seems further sustained. Knowledge cannot, therefore, be ‘objective’ in thetraditional sense of that word. The two claims – about the world and about anthropo-logical knowledge – make it increasingly uncertain what we should mean by theproposition that anthropology is based in empirical realities and that knowledge mustbe backed by evidence. While we have discussed the inherent partiality and subjectivityof anthropology over the past two decades, little effort has been made so far to under-stand the process by which anthropologists (individually and collectively) becomeconvinced of ‘being right’, or to be in the know. This is what must now be addressedfor a renewal of anthropological authority.

The concept of evidence in general connotes a strong sense of positivity, not unlike‘context’ as discussed by Johannes Fabian (1999). Anthropologists have appealed to


458




context for an accurate understanding of social events and institutions, and the claim topositivity is related to the need to establish a stable referent outside the proposed under-standing of a particular social phenomenon. This exteriority is clearly operative in theclaim that social facts must be understood ‘in context’, as if the context were indepen-dent of the scope of the actual investigation (Dilley, 1999). A similar problem holds forthe appeal to ‘evidence’ for the establishment of a particular truth claim, implying thata specific understanding of a phenomenon is confirmed by positive instances, accessibleto direct observation; these instances are then the evidence of the validity of the generalclaim. Already at this point the latent circularity destabilizes the acclaimed positive statusof evidence, as something outside of and independent of the generalization made. Thisis not a new insight; R.G. Collingwood (1978[1946]), for instance, noted that questionand evidence are correlative in the sense that facts only are evidence in relation to aparticular question (Collingwood, 1978[1946]: 178ff.; see also Chandler et al., 1994:1). Reflecting back on the previous discussion of knowledge, we might say that whateverevidence we appeal to for the explanation of a social fact, it is simultaneously evidenceof a particular field of interest, and of a particular relationship to the object.

To suggest that evidence is evidence only in relation to a particular frame of investi-gation, by itself does not necessarily question the factual nature of the evidence invoked,however. We must therefore take the question of evidence a bit further and query itsrelationship to ‘facts’. Along with an inherently positivist attitude to evidence came acommonplace that ‘facts are evidence in potentia’ (Daston, 1994: 243). This implies thatfacts and evidence are crucially distinct categories:

On their own, facts are notoriously inert – ‘angular’, ‘stubborn’ or even ‘nasty’ in theirresistance to interpretation and inference. They are robust in their existence andopaque in their meaning. Only when enlisted in the service of a claim or a conjec-ture do they become evidence, or facts with significance. (Daston, 1994)

In this view, evidence incorporates the solidity and neutrality of facts that are simply putto use in a particular argument. Contrary to this view, one could argue (with Putnam,1991) that there are no facts without value because the very identification of ‘facts’implies a particular charging of the selected phenomena, and facts cannot, therefore, bevalue neutral. The fact–evidence configuration must be revisited in its particularhistorical moment (Chandler et al., 1994: 2). At present, it seems impossible to evenspeak of ‘stubborn facts’ without implying a particular scheme of understanding towhich they do not easily lend themselves as evidence. Social facts are doubly ‘schemed’,so to speak, locally and anthropologically. They are instances of social life that areidentified as meaningful, and about which the anthropologist aims at organizing somegeneral knowledge.

In anthropology, the instances to be generalized about consist in once-occurring acts(including speech-acts), emergent meanings, and unique events (in contrast to earlierpositivist claims about empirically identifiable structures), and it is difficult to see howsocial facts of this kind can be transformed into positive evidence. The actions and eventsare real enough, of course, but as ‘instances’ they simply do not add up to evidence forthe anthropological understanding of their implications – past, present and future.‘Adding up’ does not explain how or why particular acts were undertaken, let alone why


459




they possibly made sense both to the agent and to his or her surroundings. In the wordsof Ian Hacking, truth is not explanatory – even if the style of reasoning does imply a setof self-stabilizing techniques (Hacking, 1992: 14–15).

Reasoning, however, always implies a certain amount of inference; so also foranthropological reasoning, which must be explicit of why particular instances are(perceived to be) connected in a certain way – and thereby explained. This displaces theproblem of evidence from the ontological to the epistemological domain of knowledge.The world is what we must know but not simply as it knows itself; so while the actual-ities of the world (or the field) give substance to our thoughts, it cannot at the same timebe external evidence for the connections we make. The positivity of evidence is shat-tered; there is simply not enough distance between facts and evidence for the latter toprove the former. To get it right we have to shift our attention from the collapsedrelationship between knowledge and evidence to the processes by which we know, thatis how we make connections in a fragmented world, and how these are imbued withparticular styles of reasoning. (This is not a constructionist point; even if truth claimsare linked up with particular modes of knowing, to claim that the world is constructedis still but a truth claim that begs the question of why it is constructed in a particularway. It is my view that certainly not all constructions are possible.)

In anthropology, the mode of knowing is deeply imbued with the tradition of field-work, steeping the anthropologist in a network of social relations (Hastrup, 2004b). Thiscomplicates matters of evidence further, because the once-occurring events and uniqueacts that the anthropologists register in the field at least in part are circumstantialresponses to their own presence and inquiry. This challenges another implicit quality ofevidence, namely that it must be free of human intention:

Facts fabricated as evidence, that is, to make a particular point, are therefore dis-qualified as evidence. Nature’s facts are above suspicion, because presumed free of anyintention, but many man-made facts also qualify: the blood-stained weapon foundat the scene of a murder counts as evidence as long as it was not planted there withthe intention of incriminating; the unaffected simplicity of the witness adds weightto testimony as long as it was not feigned with the intention of persuading. (Daston,1994: 244)

In science a welter of methodological precautions have been taken to thwart the (uncon-scious) intention to confirm a particular hypothesis, such as the double-blind clinicaltrial, and the fixing of statistical levels before the experiment (Daston, 1994). Withregard to the latter it is interesting to note how statistics itself has evolved in the border-land between being science and servicing science, allegedly limiting itself to the record-ing of facts (Poovey, 1994). This self-imposed limitation makes claim to a transparentrelation to the objects represented, while masking the meanings that are thereby put intoplay. ‘Largely though not exclusively an effect of the categories by which statisticalrepresentation organizes materials, these meanings are being constructed before thestatistics are compiled; they then radiate from the starkest tables. It is partly because ofsuch statistical representation – even if it is nowhere acknowledged – that theory andlegislation can be generated from numbers’ (Poovey, 1994: 420). While this may notcome as a surprise to category-conscious anthropologists the implications are profound,


460




because they touch upon the problem of intention and evidence that has just been raised.Statistics not only depends on the unfigured, because its general estimates are inferredfrom more limited counts (Poovey, 1994), it also importantly derives its significancefrom the unconfigured, in other words that which lies outside the scope of the intendedgeneralization – beyond the categories that count.

There is no way for statistics to epistemologically acknowledge its ‘excess’ meaning;in anthropology, by contrast, the phenomenological bent always contributed to anawareness of the historical surplus of any moment, any event. By engaging with theworld under study in fieldwork, anthropologists perceive and realize the surplus ofexperience that qualifies social life. Not all experience is reducible to knowledge ( Jackson,1996: 3); phrased otherwise, anthropological fieldwork discloses the fact that there isalways a historical surplus of events, actions and thoughts that may linger withoutnecessarily contributing to the larger order as perceived, but providing possible sites ofresistance or sources for new historical turns. While statistical evidence thus easilybecomes a normative argument, this is not so for anthropology, where evidence cannoteasily be ‘added up’, as I suggested earlier. The (implicit) intentionalities of the two disci-plines are simply very different – as are their relative authority in present-day westernsociety, generally attributing more ‘reality’ to numbers than to countless experiences.

It will be clear by now that the question of evidence is far more complex than theclaim to positivity suggests, because it is enfolded within the relational nature ofanthropological knowledge that – epistemologically – precludes the use of evidence asan independent measure of validity. Another complication is found in the fact that somuch evidence in anthropology is circumstantial or inferential, and relates to sensations,silences, deceptions, and moods. These sensations are not external to categories butinform them deeply; as Jon Mitchell has shown, for instance, feelings are all-importantin the analysis of belief (1997). Space does not permit me to discuss particular cases, soI shall just invite readers to speculate on the nature of ‘evidence’ and its relation to suchsocial phenomena as values, jokes, lies, and taboos – not to mention roots, identities,and democracies. We do not hesitate using such notions and attributing reality to them,but we have failed to ask how (or if ) they link up with facts outside the situation inwhich they are perceived as true by the anthropologist – as opposed to simply being sign-posted as such by the joker, the liar or the prime minister. The question still is how wemay acknowledge the ‘rightness’ of what cannot be empirical knowledge in conventionalpositivist terms. In order to come closer to a temporary answer, I shall introduce ahistorical example.

HISTORICAL ANTHROPOLOGY: IDENTIFYING CAUSESIn the humanities, the historical disciplines have probably been more explicitly concernedabout evidence than most. I already referred to Collingwood’s early piece on ‘HistoricalEvidence’ (incorporated in The Idea of History, 1978[1946]), which alerted historians tothe mutuality of question and evidence. Since (and before) then, historians have beenengaged in a more or less enlightened discussion of ‘historical sources’ and their reliability.Carlo Ginzburg (1994) has suggested that the discussion of evidence in history is closelylinked to a notion of proof that draws on an implicit parallel between the historian andthe judge. While their objectives are vastly different, they both have to look for clues,leads, distortions, and discrepancies in the material collected to support their case.


461




If justice is the concern of the judge, accurate reconstruction of past events is theambition of the historian, or so it was held. What easily slips by the historian’s notionof reconstruction is the impossibility to ever depict an event in its totality of impli-cations; it is precisely the limitation of ‘evidence’ that makes the illusion of wholenesspossible. Thus once again we are faced with a situation where the mode of knowing(through transmitted sources) greatly influences what can become knowledge. Indeed,the notion of ‘history’ itself imposes a particular view of what should be known. Asdiscussed elsewhere (Hastrup, 1992a), scholars (including anthropologists) have largelybeen victims of a Eurocentric view of history as linear and continuous. Consequently,our sense of history has been defined as a particular mode of consciousness, whichassumes social change to be homogeneously progressive (Lévi-Strauss, 1962); in anthro-pology ‘the others’ seem to have been excluded from ‘our’ history and placed in adifferent, non-progressive time (Fabian, 1983).

The narration of history makes the linear mode of historical consciousness manifest;the plot is constructed as a succession of instances, bound together by a logic of causa-tion, sometimes in the form of a teleology. Indeed, the entire construction of narrativeis bound to and dependent upon the construction of time and sequence and bothdepend on experience for their construction: ‘time becomes human to the extent that itis articulated through a narrative mode, and narrative attains its full meaning when itbecomes a condition of temporal existence’ (Ricoeur, 1984: 52). Narrative emplotmentbrings together agents and circumstance and connects means and ends in unexpectedways (Ricoeur, 1984: 65).

The linear narration of history is linked to an idea of ‘temporal causation’ that hasdominated the historical discipline, identifying the cause of an event in the most recentand most extraordinary precedent. This notion is not sustainable, because it leaves outwhat seems ‘ordinary’ and absolutely necessary at the time for the thing to happen(Bloch, 1979: 191). Furthermore, it does not question ‘time’ itself. It remains an openquestion how much time one can allow between cause and effect and still speak of acausal process. Somehow, for the idea of temporal causation to be convincing, the timebetween cause and effect must be ‘filled out’, so to speak, by the causal process. In humanhistory it is simply not possible to provide positive evidence for this proposition. To linkevents we must resort to a sense of duration that can only be established imaginatively– and proposed narratively. Causation is therefore embedded in duration.

My thoughts on this matter derive from my study of Icelandic history (Hastrup, 1985,1990, 1998a) from which I shall give a condensed example. In the briefest possibleterms, the period from 1400–1800 in Iceland was a time of recurrent crises – social,economical, demographic, environmental, climatic, you name it. In each of thesedomains there was a remarkable backsliding in comparison to the earlier period of thesettlements and the high Middle Ages, from whence the famous Icelandic sagas surviveas a testimony to an original ‘free’ society, astute statesmanship and legal sophistication– not to mention the literary achievement itself. By this standard the devolution wasremarkable; for each of the domains just mentioned we might name a cause – bypointing to what was the most spectacular incident in the course of events. I am thinkingof events like the Black Death, volcanic eruptions, the introduction of a Danish trademonopoly and so forth. In naming such causes – all of which will be familiar tohistorians as the traditional explanations of the decline – we have not answered the


462




question of causation however. We may have identified significant parts of the empiri-cal reality, but we have explained very little beyond immediate local experience. We havebeen trapped in the circularity of positive evidence and in a narrative logic of causationthat both relate only to the succession of events but cannot account for the links betweenthem. Truth does not equal explanation; truth is not a predicate and adds nothing tothe subject (Hacking, 1992: 14).

Such links require an attention to the ordinary, the durable, including (in this case)the sense of Icelandicness that prevailed and which prompted particular responses to theevents. One such response in Iceland was to concentrate labour in farming at the expenseof fishing whenever crisis struck, which made the economy even more vulnerable andprecipitated a negative spiral. We cannot ask contemporary Icelanders about theirreasons for responding to circumstance as they did, gradually cutting themselves off frompotential sources of supplementary income, and relegating more and more people to thecategory of non-humans (for their poverty, lack of permanent dwelling and so on). Whatwe can do is to infer from a wide range of sources that Icelandicness itself was associ-ated with a settled farming life, and to protect that definition, the structurally dominantIcelandic farmers drew an ever-diminishing circle around the ‘Icelanders’ proper.

The details are many and the material is very rich, but space only allows me to saythat alongside the sequential historical narrative of a particular succession of events, therewas a strong sense of duration – a duration that filled out the entire temporal process ofthe decline. We are allowed, therefore, to suggest that while the combination of unhappyfactors certainly hit the North Atlantic community hard, they neither individually norin combination caused the crises on their own. To explain the course of history and notsimply point out the obvious, the enduring model of a society inhabited by free andequal farmers must be taken into account. This model is inferred from a variety of eventsand acts, and as such it is a theoretical construct. The point is, that at least part of theexplanation for the historical development must be found outside the positive events ina proposed connection between them that relates to a durable sense of Icelandicness.This sense implied particular responses to events – urging the Icelanders to recurrentlyrecast themselves in terms of past values and ideas. I have suggested the notion ofUchronia for this implicit appeal to ‘a history out of time’ (Hastrup, 1990, 1992b). Theevidence for an Icelandic Uchronia is circumstantial rather than empirical, and is linkedup with a particular question of causation, but it is none the less real for that.

The general point is that a sense of duration always embraces whatever notion wemight have of sequence, because events only register as such within a frame that outlaststhem. ‘Evidence’ in the inherited, positive, sense of the word attaches itself to eventsrather that duration; duration has to be inferred from certain regularities in the socialabout which individual people caught up in a struggle for survival cannot themselvestheorize. From the point of view of the acting individual, it is impossible to be insideand outside of a particular history at the same time. One cannot be absorbed in onetime, in one vision of history, and act in another.

To reiterate, causes cannot be identified empirically or ontologically, but they maysuggest themselves in relation to particular schemes of understanding, or particular epis-temologies; this applies equally to local (or lay) views of causation and to historians’ (orother scientists’) views. The identification of causes beyond the most simple pointingout of particular events (being truths rather than explanations) can therefore not be


463




backed by objective, measurable evidence; the total historical development itself is asymptom of something more complex – and it becomes evidence only by virtue of itsgiving rise to a particular theoretical understanding of the relationship between actionand history.

This point in some ways echoes the conclusion reached by Hume in the mid-18thcentury, when he discussed causality as a construction upon past experience and claimedthat an idea of causation cannot be drawn directly from the external world, but mustpass by the ruling perceptions of that world. What is more, Hume identified imagin-ation as the principle which ‘makes us reason from causes to effects and it is the sameprinciple which convinces us of the continued existence of external objects’ (Hume,1976: 266). This leaves us with the classical ‘problem of induction’ and takes us furtherinto the discussion of the distinctly anthropological grounding of knowledge.

FIELD EXPERIENCE: MAKING CONNECTIONSSince the pioneering work of Émile Durkheim, anthropologists have been preoccupiedwith the tension between the whole and the parts, or between society and individuals.The analytical focus has shifted and either the whole or the part has been deemedlogically prior, but the tension itself never goes away. Recently, it has been suggested that‘culture’ simply provides a set of shared images, values and modes of speech to whichindividuals may attach their individual understandings (Cohen, 1994). Within anysocial field both the shared images and the diversity of individual understandings presentthemselves as facts. The question is how we can establish the connection between theindividual act and the larger scheme of things without falling into a determinist trap.

Given the particular anthropological mode of knowing, the initial answer to thequestion is that we generalize from our own experience in the field, not from someabstract ideas that are then backed by evidence. The distinctly performative mode ofknowing implied in fieldwork – as opposed to the informative and observational(Fabian, 1990: 3ff.) – immediately links understanding to participation, and evidenceto experience. In the field, anthropologists engage in social relationships in order to feeltheir nature and directive force. To understand is to acknowledge one’s own partici-pation, and to allow oneself to be ‘caught up in the series of events that constitute sociallife, where there is no objective truth, but simply potentially exclusive versions of thetruth that together constitute the event’ (Jenkins, 1994: 443). This urges us to take acloser look upon the notion of experience and its implications of authenticity andconclusiveness.

Part of my own fieldwork in Iceland was among farmers, and on the farm where Ilived, I was given the role of milkmaid – among others (see Hastrup, 1998a). For acouple of months I helped milking and tending 30 cows – not very skilfully at first.Taking the cattle to the pasture was my major problem in this rugged and unboundedlandscape. I had to keep the herd together across large and open stretches of more orless rocky ground before arriving at the grazing slot for the day, bounded by a mobileelectric fence. Once, the whole herd went madly astray, and it took a couple of menseveral hours on horseback to collect them again. I was embarrassed, of course, but byway of consolation one of the farmers smilingly told me that ‘cows are stupid’. The pointis not to exhibit my remarkable lack of skill, but to point to its source. It is not simplythat I had been working at a desk for most of my life. More importantly, I could not act


464




adequately, when I still saw the cows as a herd, even a category (a view that was perceivedby the farmer, and expressed in his consoling words to me), and myself as an anthro-pologist performing as milkmaid. I could only succeed when I truly began living thecharacter. That is, when I began to see the cows as named individuals with distinct behav-ioural dispositions, and when I gave up my resistance to yelling at them on my way tothe pasture, in what at first seemed to be an almost obscenely loud and distorted toneof voice, and to use my stick.

In so far as fieldwork implies actual presence in the social world, the experience isrelated to living our part (Hastrup, forthcoming a). This ‘part’ is very much a part allottedto us by the others (Hastrup, 1987); not all parts are available. From the authoritativesubject position of the researcher, the ethnographer is intermittently transferred into anobject position in the world of the subjects studied. The anthropologist in the fieldengages the world as a ‘double agent’, being both a trained researcher and a character inthe local drama (Hastrup, 1998b). It is from this position that the anthropologist beginsto experience the complexities of an unfamiliar everyday life; knowledge becomes incor-porated (Hastrup, 1994). The position entails a new question about the subject ofexperience. It is not simply the anthropologist in the First Person, nor is it a (putatively‘native’) Third Person, playing an allotted part in an exotic drama. It is a peculiar char-acter embodying both, and emerging in the context of the situation, thereby remindingus that even subjects must be historicized. Therefore, experience cannot be externalevidence of a particular situation, because experience cannot be attributed to an indi-vidual who stands outside of the situation that the experience is evidence of.

Another example will take us further into this. Like other anthropologists I was keenlyinterested in oral tradition, and not least in the ‘hidden people’ (huldufólk) – an elf-likepeople that lived invisibly in a parallel world to the world of humans. I asked aboutthem, and got the inevitable answer that they were something that people of the pastbelieved in. Together we rehearsed some well-known legends and folktales. Only laterdid I realize that both the belief and the old days were still very much present. I kepthearing half-joking references to the ‘hidden people’ stealing the tools and misplacingthe utensils they had borrowed, and the landscape was heavily marked by their presencein place-names. It took a particular experience for me to realize that the jokes veryliterally were directed at me, not out of spite at all but out of respect for my limitedknowledge. Participating in a venture into the mountains after stray sheep I was onceleft on a mountain ledge in care of a ewe that had been recovered. Instinctively, I woundthe rope tightly around my hand, but was warned not to do so, because it was essentialalso to be able to let go, should the animal choose to jump over the cliff. This was animportant piece of information, and it certainly contributed to a sense of foreboding asI sat there, left to myself and to the wide-eyed animal, and – moreover – soon becameswallowed up in one of the unpredictable mists that swept the coastland mountains. Inthe mist, my fear of being lost was further enforced by the intense feeling of nebulousfigures emerging out of it, and ‘reminding’ me of countless stories of shepherdesses beingswept off their feet by huldumenn (men of the ‘hidden people’), apparently asserting theirpresence even now.

Whatever one would make of this experience, it made me pose a different sort ofquestion next time the issue of the hidden people came up in conversation. I simplyasked when they had last seen the hidden people in the vicinity of the farm, and after


465




some internal debate on relative chronologies and various sightings a date some yearsback was suggested. The point is, that the new framing of the question had produced aqualitatively different answer; I was no longer asking for information about a category,but enquiring about significant experience. My phrasing implied an acknowledgementof another way of looking at things, and henceforth the farmers could talk to me abouttheir vision of the world in terms that went beyond mere information. To establish atrue relationship the parties must be present in the same space; in fieldwork this has tobe theirs – if there is any point to participation. The general point is that living a particu-lar social field implies a merging of action and awareness (Hastrup, 2004a). This mergingis the basis for the self-evidence of incorporated ‘local’ knowledge – also in everyday life– and a prerequisite of any skill or practical competence, including the skill at posingmeaningful questions. The point is to get away from the dualism of thought and actionin recognition of the fact that knowledge is practical, and that theoretical or abstractknowledge is a special case of this (Jenkins, 1994: 442). This is a correlate to claimingthat there is no opposition between practical (material) experience and its theoretical(linguistic) rendering; they are deeply implicated in one another.

This is where we begin to see that the qualitative nature of anthropology reflects thequality of social life itself; the situation of the fieldworker is characteristic of theconditions being studied (Jenkins, 1994: 442). This is the reason why fieldwork is a validway of gaining knowledge about other people, even if it cannot be backed by positiveevidence in the old sense of the term. I cannot point to a hulduma!r and claim hispresence as evidence of anything outside of the experiential situation; conversely, theexperience is not evidence of the factuality of huldumenn. ‘Experience’ cannot be foun-dational; firstly, because one is naturally implicated in the situation in which the experi-ence occurs, and secondly, the experience is imbued with interpretation – according toavailable schemes of understanding and categories. As these schemes shift so do experi-ences; as indicated earlier so, even, do subjectivities.

By acknowledging that our own participation in the world under study is a distinct(anthropological) avenue towards understanding, the search for independent or externalevidence breaks down. The connections we make are inferred from our being implicatedin them. This, again, locates ‘rightness’ in an epistemological awareness rather than inontological certainty. It also shifts the objective of generalization from being (primarily)an identification of shared systems of meaning to the processes by which meanings areestablished, challenged and altered; that is the shift from horizontal to vertical general-ization proposed earlier. In other words, stampeding cows and nebulous figures in mistymountains may not have taught me a lot about Icelandic ‘culture’ as such, but they didteach me about the trap inherent in thinking in categories, rather than situations, andby implication I learnt something about a particular perception of the environment thatcontributed to a demarcation of Iceland as a place apart (Hastrup, 1998a).

We have come a long way from the anthropologist portrayed by Carlo Ginsburg(1989) as akin to the inquisitor. He discusses how the rich evidence for the thoughts ofordinary people brought about by the inquisition is of necessity distorted by the presenceand pressure of the trial itself; the inquisitor’s urge for truth made the defendants moreor less spontaneously reproduce the stereotypes of the inquisition itself. And Ginzburggoes on: ‘Similarly, the comparison between inquisitorial trials and anthropological fieldnotes could have, from the historian’s point of view, a negative implication: the pressure


466




of those long-ago anthropologists would be so obtrusive as to prevent us from knowingthe beliefs and thoughts of the unhappy natives brought before them’ (Ginzburg, 1989:158–9). He here attributes anthropology with the ambition to find the truth about othercultures (as did anthropologists of yesteryear) – and sees both unhappy natives and un-reliable anthropologists as the result.

Anthropology is never better than the individuals who practice it, as Geertz hasobserved. If thinking is a profoundly ‘moral’ act, in the field you have to think and liveat the same time (Geertz, 2000: 21ff.). You live from moment to moment, seeking tounderstand it, all the while being trapped in the eventness of being. The irony is that allthe while you are engaging the local social order (of whatever scale, and wherever) andliving your part, you come to embody a whole (a plot) that you can only partly under-stand, and certainly only describe through inference. Through our actions we stand outas particular ‘characters’ (Hastrup, forthcoming a, forthcoming b). For any social agent,the connection between the ‘plot’ (whether tending some cows or ruling a state) and theindividual act is embodied; conversely, the wholeness of the plot is present in the indi-vidual action.

As Jean-Pierre Vernant has said about the world of Aristotle’s dramas, the individualwho commits a deed is also its victim:

The action does not emanate from the agent as from its source; rather it envelopshim and carries him away, swallowing him up in a power that must perforce bebeyond him since it extends, both spatially and temporally, far beyond his ownperson. The agent is caught in the action. He is not its author; he remains includedwithin it. (Vernant, 1992: 44)

Again, while we may be able to establish the facticity of particular events and actions(by observation), they are not direct evidence for the larger plot that we may infer fromthem. Inferring (and imagining) is an integral part of thinking in general, and for theanthropologist the challenge is to infer alternative schemes of inference – or alternativeprinciples of imagination, in Hume’s terms, as briefly mentioned earlier.

To summarize the argument of this section: Just like historical ‘sources’ can never belocated outside of their own context of production and use, and like historical eventscannot be by themselves evidence of a history that connects them, so experience cannotbe taken as a foundational concept, explaining social relationships and providing externalevidence for connections made. Experiences must themselves be explained because in asense they are already imbued with interpretation; they are neither neutral nor predis-cursive (Scott, 1994: 387). The anthropological experience in the field, therefore, is notdirect evidence of a particular ‘culture’, but of the ways in which particular modes ofaction come to present themselves as ‘ordinary’ and how they are liable to individualcontestation. The ordinariness points not to a stable and essential culture, but to amoving framework of action, a plot that gradually shifts while people are playing theirpart.

The connections that the anthropologist makes are not so much backed by an experi-ence of culture as by an experience of the contingency of frames within which every-body plays his or her part. This cannot simply be observed in the everyday, becausepeople and frames move simultaneously in ordinary life; similarly, change cannot be


467




measured while it happens. It takes the effort of a deliberate exposure to ‘other frames’,as implied in the tradition of fieldwork and in the particular anthropological attentionto be able to sort it out (Hastrup, 2004b). Lest I be misunderstood, I want to stress thatthis is no disclaimer of an anthropology in one’s own society; I am speaking of a delib-erate alienation from the world under study in order to understand it as it cannot under-stand itself. This is where we can substantiate the anthropological ‘effort’ atunderstanding (signposted in the initial discussion of knowledge earlier in this article),and where we can make new claims to the necessity of fieldwork even though we cannotclaim experience to be foundational.

ANTHROPOLOGICAL KNOWLEDGE: GETTING IT RIGHTWith the acknowledgement of knowledge as relational, both historical events and socialexperience have lost their immediate status as positive evidence of the connectionsbetween events, actions and experiences. Such connections are inferred by historians andanthropologists aiming at a kind of explanation beyond the truth of the events them-selves. Anthropological knowledge, then, is not simply knowledge about particularevents, practices and ideas, but about the processes by which these come to appearmeaningful, perhaps inevitable or mandatory, possibly contestable or even mad. Theseprocesses can be seen as contributions to the over-all process of objectifying the socialin so many ways. To get at these the anthropologist must engage the world under study,but must also at the same time realize that the anthropological object is emergent.Because knowledge is gradually incorporated, the object has no fixed ontological status,be it as a culture, a society or a community. It emerges in consequence of a particularintervention and analytical scope. Conversely, the anthropological subject does not standoutside of the situation under study, and at both levels we can see how the particularrelation to the object of necessity bends back into the object itself. This demands aparticular sensitivity to the epistemology of anthropology, if its authority shall beasserted.

Anthropological authority is not to be found by appeal to external, factual evidence– as indicated in my initial discussion of the positivity of evidence. If, as I claim foranthropology, our relation to the object is already installed as part of the object whenwe begin to understand it, ‘evidence’ cannot be disengaged from the objective of theinvestigation. It is tempting, perhaps, to conclude that the absence of positive evidencefor a particular anthropological interpretation makes it impossible to assess its ‘right-ness’. There seems to be no standard for validating or questioning a particular point.This was very much implied by the radical post-modernist stand, disclaiming any notionof truth and therefore in principle making all ‘stories’ equal.

This is unwarranted. Not only does it entail a narrative irresponsibility (that ought tobe foreign to scholarship), it also entails a falling back into positivist thinking from whichpost-modernism – if anything – ought to have freed anthropology. The point of anthro-pology is not to tell the world as it is (which would be practically impossible) but tointerpret it and to suggest possible (theoretical) connections within it as perceived andinferred from being in touch with a world that cannot be taken for granted – unlike thehome world. By definition, the ‘home world’ is where frames and events are seamlesslyand imperceptibly fused.

While we may have endorsed the demise of the grand narratives and their implied


468




truth claims, we are still in need of narrative imagination (Kearney, 1998: 241ff.). Theobjective of the anthropological narrative (of whatever scale) is to provide a mode ofimagining how individual actions and collective illusions are interlinked, and how theyare framed by an implicit sense of a common good. To explain how the world works isnot to explain it away but to make new connections between social ‘facts’ that mayprovide unprecedented insight into the workings of social worlds in general. One suchinsight concerns the excess of social experience – the historical surplus – at any point oftime, that is experience which is not captured by current categories, and which pointsto alternative ways of seeing things and acting upon them – and hence to possible sitesof social resistance or creativity, as the case might be. In their field practice, anthro-pologists experience both the contingency of the given, and its efficiency.

The absence of positive evidence for the durability and connectedness of social factsdoes not entail a narrative irresponsibility. Quite the contrary, the profoundly perfor-mative and relational mode of knowing in anthropology implicates us deeply in a narra-tive ethics that is the source of anthropological authority, and gives force to the argumentthat connects the ‘true’ in new ways (see Hastrup, 1995: 162ff.). Not all stories makesense, because the social ‘facts’ – such as they are – do not connect in any number ofways; anthropology is ‘realist’ in the sense of having to take perceived realities seriously.Because anthropological knowledge is relational rather than objective, the latent posi-tivity of evidence should be revised and recast with a view to epistemological consider-ations – also on the process by which cognitive economizing compels us to dump ourtemporary understanding as objective knowledge.

To speak of a narrative ethics, however, is not simply to acknowledge that connec-tions between social facts and larger frames cannot be ‘constructed’ at random; it is alsoto acknowledge that the identification of facts is imbued with value (Putnam, 1991).While Hume’s law – that one cannot infer an ‘ought’ from an ‘is’, or that ethics is notabout ‘matters of fact’ – is still held to be generally valid, the fact/value dichotomy isuntenable when it comes to knowledge, because (as the pragmatists have it) knowledgeof facts is knowledge of values (see Putnam, 2002: passim). Practising anthropologyimplies a ‘using’ of other people’s understanding to further an anthropological under-standing that is narratively mediated. In writing, anthropologists make connections andsort out hierarchies of significance that cannot bypass local social knowledge even whiletranscending it. The ethical demand is to ‘get it right’, not in any ontological sense, butin being true to the world under study and to the epistemological premises of anthro-pology. Rightness itself is a value, an ethical imperative imposed upon the narrativeimagination of social and historical relations – including causal relations. The narrativeethics that I am propounding is an ‘ethics without ontology’ (Putnam, 2004), butcertainly not without reality or power. In discussing ‘the narrative imperative’, MichaelJackson suggests that ‘storytelling is a vital human strategy for sustaining a sense ofagency in the face of disempowering circumstances’ (2002: 15). I would want to takethis a bit further and suggest that the narrative imperative in anthropology is to alert theworld to the force of the everyday, to show the historical surplus of the moment, andpoint to sites of resistance in the given.

‘Getting it right’ is backed by anthropologists being in touch with reality – not bystanding outside it looking for evidence – and it is further sustained by a narrativeimagination that figures out how parts and wholes are constructed and how individual


469




acts and communal images are both mutual preconditions and challenges. ‘Figuring itout’ is to configure what at base is defiant of configuration and to do so respectfully isto acknowledge the mutual implication of epistemology and ethics.

AcknowledgementsA shorter version of this article was given at the AAA meeting in Chicago, November2003, in the session on ‘Evidence and its culture’ sponsored by the Wenner-Gren Foun-dation. I want to thank Richard Fox for his invitation and co-speakers and audience fora stimulating and lively event.

ReferencesBarth, Fredrik (2002) ‘An Anthropology of Knowledge’, Current Anthropology 43(1):

1–18.Bloch, Marc (1979) The Historian’s Craft. Manchester: Manchester University Press.Bourdieu, Pierre (2000) Pascalian Mediations. Cambridge: Polity Press.Chandler, James, Arnold I. Davidson and Harry Harootunian (1994) ‘Introduction’,

in James Chandler, Arnold I. Davidson, and Harry Harootunian (eds) Questions ofEvidence. Proof, Practice and Persuasion across the Disciplines, pp. 1–8. Chicago, IL:University of Chicago Press.

Cohen, Anthony (1994) Self Consciousness. An Alternative Anthropology of Identity.London: Routledge.

Collingwood, R.G. (1978[1946]) ‘Historical Evidence’, in R.G. Collingwood The Ideaof History, pp. 249–82. Oxford: Oxford University Press.

Crick, Malcolm (1982) ‘Anthropological Field Research, Meaning Creation andKnowledge Construction’, in David Parkin (ed.) Semantic Anthropology, pp. 15–37.London: Academic Press.

Daston, Lorraine (1994) ‘Marvelous Facts and Miraculous Evidence in Early ModernEurope’, in James Chandler, Arnold I. Davidson, and Harry Harootunian (eds)Questions of Evidence. Proof, Practice and Persuasion across the Disciplines,pp. 243–74. Chicago, IL: University of Chicago Press.

Dilley, Roy (1999) ‘Introduction’, in Roy Dilley (ed.) The Problem of Context,pp. 1–46. Oxford: Berghahn.

Durkheim, Émile (1915) Structures élementaires de la vie religieuse, Paris. (Later edition[1976] The Elementary Forms of Religious Life. London: Allen and Unwin.)

Evans-Pritchard, Edward E. (1936) Witchcraft, Oracles and Magic among the Azande.London: Routledge and Kegan Paul.

Fabian, Johannes (1983) Time and the Other. How Anthropology Makes its Object. NewYork: Columbia University Press.

Fabian, Johannes (1990) Power and Performance: Ethnographic Explorations throughProverbial Wisdom and Theater in Shaba, Zaire. Madison: University of WisconsinPress.

Fabian, Johannes (1999) ‘Misunderstanding and the Perils of Context’, in Roy Dilley(ed.) The Problem of Context, pp. 85–104. Oxford: Berghahn.

Geertz, Clifford (2000[1968]) ‘Thinking as a Moral Act’, in Clifford Geertz AvailableLight. Anthropological Reflections on Philosophical Topics, pp. 21–41. Princeton, NJ:Princeton University Press.


470




Ginzburg, Carlo (1989) ‘The Inquisitor as Anthropologist’, in Carlo Ginzburg Clues,Myths, and the Historical Method (trans. John and Anne C. Tedeschi), pp. 156–64.Baltimore, MD: Johns Hopkins University Press.

Ginzburg, Carlo (1994) ‘Checking the Evidence: The Judge and the Historian’, inJames Chandler, Arnold I. Davidson, and Harry Harootunian (eds) Questions ofEvidence. Proof, Practice and Persuasion across the Disciplines, pp. 290–303. Chicago,IL: University of Chicago Press.

Hacking, Ian (1992) ‘ “Style” for Historians and Philosophers’, Studies in the Historyand Philosophy of Science 23(1): 1–20.

Hastrup, Kirsten (1985) Culture and History in Medieval Iceland. An AnthropologicalAnalysis of Structure and Change. Oxford: Clarendon.

Hastrup, Kirsten (1987) ‘Fieldwork among Friends’, in Anthony Jackson (ed.)Anthropology at Home, pp. 94–108. London: Routledge.

Hastrup, Kirsten (1990) Nature and Policy in Iceland 1400–1800. An AnthropologicalAnalysis of History and Mentality. Oxford: Clarendon.

Hastrup, Kirsten (1992a) ‘Introduction’, in Kirsten Hastrup (ed.) Other Histories,pp. 1–13. London: Routledge.

Hastrup, Kirsten (1992b) ‘Uchronia and the Two Histories of Iceland’, in KirstenHastrup (ed.) Other Histories, pp. 102–20. London: Routledge

Hastrup, Kirsten (1994) ‘Anthropological Knowledge Incorporated’, in KirstenHastrup and Peter Hervik (eds) Social Experience and Anthropological Knowledge,pp. 224–40. London: Routledge.

Hastrup, Kirsten (1995) A Passage to Anthropology. Between Experience and Theory.London: Routledge.

Hastrup, Kirsten (1998a) A Place Apart. An Anthropological Study of the Icelandic World.Oxford: Clarendon.

Hastrup, Kirsten (1998b) ‘Theatre as a Site of Passage: Some Reflections on the Magicof Acting’, in Felicia Hughes-Freeland (ed.) Ritual, Performance and Media,pp. 29–45. London: Routledge.

Hastrup, Kirsten (2004a) Action. Anthropology in the Company of Shakespeare.Copenhagen: Museum Tusculanum Press (University of Copenhagen).

Hastrup, Kirsten (2004b) ‘The Anthropological Attention and the Pertinence of theField’, Social Anthropology (in press).

Hastrup, Kirsten (forthcoming a) ‘Playing One’s Part: The Imaginative Framework ofAgency’, to appear in Stuart MacLean and Steve Coleman (eds) EngagingImagination. Dublin.

Hastrup, Kirsten (forthcoming b) ‘Identity at Play: Individuals, Characters andTheatres of Action’, in Deborah Bryceson, Judith Okely, and Jonathan Webber(eds) Fashioning Identities and Weaving Networks: Gender and Ethnicity in a Cross-Cultural Context. Oxford.

Hollis, Martin and Steven Lukes, eds (1982) Rationality and Relativism. Oxford:Blackwell.

Hume, David (1976) A Treatise of Human Nature. Oxford: Oxford University Press.Jackson, Michael (1996) ‘Introduction’, in Michael Jackson (ed.) Things as They Are.

New Directions in Phenomenological Anthropology, pp. 1–50. Bloomington:University of Indiana Press.


471




Jackson, Michael (2002) The Politics of Storytelling. Violence, Transgression andIntersubjectivity. Copenhagen: Museum Tusculanum Press (University ofCopehagen).

Jenkins, Timothy D. (1994) ‘Fieldwork and the Perception of Everyday Life’, Man 29:433–55.

Kearney, Richard (1998) Poetics of Imagining. Modern to Post-modern. Edinburgh:Edinburgh University Press.

Lévi-Strauss, Claude (1962) La pensée sauvage. Paris: Plon.Mitchell, Jon (1997) ‘A Moment with Christ: The Importance of Feelings in the

Analysis of Belief ’, Journal of the Royal Anthropological Institute 3: 79–94.Poovey, Mary (1994) ‘Figures of Arithmetic, Figures of Speech: The Discourse of

Statistics’, in James Chandler, Arnold I. Davidson, and Harry Harootunian (eds)Questions of Evidence. Proof, Practice and Persuasion across the Disciplines,pp. 401–21. Chicago, IL: University of Chicago Press.

Putnam, Hilary (1991) ‘Beyond the Fact/Value Dichotomy’, in Hilary Putnam Realismwith a Human Face, pp. 135–41. Cambridge, MA: Harvard University Press.

Putnam, Hilary (2002) The Collapse of the Fact/Value Dichotomy and Other Essays.Cambridge, MA: Harvard University Press.

Putnam, Hilary (2004) Ethics without Ontology, Cambridge, MA: Harvard UniversityPress.

Ricoeur, Paul (1984) Time and Narrative, Vol I. Chicago, IL: University of ChicagoPress.

Schmidt, Lars-Henrik (1991) Det Videnskabelige Perspektiv. København: AkademiskForlag.

Scott, Joan W. (1994) ‘Experience as Evidence’, in James Chandler, Arnold I.Davidson and Harry Harootunian (eds) Questions of Evidence. Proof, Practice andPersuasion across the Disciplines, pp. 363–81. Chicago, IL: University of ChicagoPress.

Strathern, Marilyn (1996) ‘Cutting the Network’, Journal of the Royal AnthropologicalInstitute 2: 517–35.

Vernant, Jean-Pierre (1992) ‘Myth and Tragedy’, in Amélie O. Rorty (ed.) Essays onAristotle’s Poetics, pp. 33–50. Princeton, NJ: Princeton University Press.

Wilson, Bryan, ed. (1974) Rationality. Oxford: Blackwell.

KIRSTEN HASTRUP is Professor of Anthropology at the University of Copenhagen. She has done extensive

research on Icelandic history and society, on theatre and on human rights, and she is the author and editor of

numerous books in all these fields. Her principal interest at the moment is the theoretical and epistemological

underpinning – and potential – of anthropology as a distinct discipline, as foreshadowed also in a couple

of earlier books. Address: Institute of Anthropology, University of Copenhagen, Frederiksholms Kanal 4,

DK-1220 Copenhagen K, Denmark. [email: [email protected]]


472




hastrup, kirsten - getting it right

Documents

museum tusculanum press

harvard university press

copenhagen university library

xed ontological status

necessity bends back

princeton university press

oxford university press

social life